conversation-analyzer

Analyzing Claude Code conversation history requires local access to the `~/.claude/history.jsonl` file. The Skill executes a Python script to parse this data, categorizing request types, tracking active projects, and identifying error keywords across your last 200 conversations to reveal workflow patterns.

Yes, you can identify repetitive tasks and common mistakes by scanning conversation history for repeated fixes, merge conflicts, and vague requests. The analysis detects multi-step tasks executed without planning, categorizing these occurrences to highlight areas where workflow optimization or task automation is needed.

Parsing conversation history for workflow optimization requires local access to the `~/.claude/history.jsonl` file and a Python environment with standard libraries like `json`, `re`, `collections`, `datetime`, and `pathlib`. The script processes up to 200 conversations to extract complexity indicators and usage patterns.

Conversation analysis does not cover your entire history; it scopes exclusively to the last 200 Claude Code conversations. This limit ensures efficient data parsing and relevant pattern recognition for request types, error keywords, and workflow complexity indicators without overwhelming processing.

Finding automation opportunities involves analyzing conversation history to detect repeated fixes and multi-step tasks lacking prior planning. The Skill categorizes these repetitive actions and suggests areas where new skills could be created to streamline your development workflow.

Vague requests are flagged during conversation history analysis because they indicate workflow inefficiencies and multi-step tasks executed without planning. Identifying these unclear prompts helps pinpoint areas where improving request specificity could optimize your Claude Code interactions and reduce complexity.